摘要
在数据中心网络中,实时虚拟机迁移有利于实现网络的优化和管理目标。为了有效的进行虚拟机迁移,减少总迁移时间,提升服务性能,提出基于分组的启发式算法(grouping-based heuristic algorithm,GBHA)。算法通过计算每个分组的虚拟机对其他虚拟机迁移时间的累积影响,及时更新有限的网络带宽等资源约束,找到最优的迁移方案。算法在保证依赖关系和性能要求的前提下,解决了大规模迁移触发时的迁移规划问题,减少虚拟机的总迁移时间。实验结果表明,与CQNCR算法和HACE算法相比,该算法能够将虚拟机的总迁移时间分别降低44.1%和26.5%,有效地提高迁移性能。
Live virtual machine( VM) migration is an effective technique to achieve network optimization and management goals in data center networks. For making an effective VM migration planning to reduce the total migration time and improve the service performance,a grouping-based heuristic algorithm( GBHA) is proposed in this paper. The proposed algorithm finds the optimal migration plan by calculating accumulated impact of each group ’s VMs on the migration time of other VMs and timely updating the available network bandwidth resource and so on. The algorithm solves migration planning problem when the massive migrations are triggered and reduces the total migration time of VMs under the premise of ensuring the dependence and performance requirements. The experiments show that the algorithm can reduce total migration time of VMs by up to 44. 1% and 26. 5% compared with the CQNCR and the HACE algorithm,respectively,and the migration performance is improved effectively.
出处
《电子测量与仪器学报》
CSCD
北大核心
2018年第2期119-127,共9页
Journal of Electronic Measurement and Instrumentation
基金
国家重大科学仪器设备开发专项(2013YQ030595)资助
关键词
软件定义网络
数据中心
虚拟机迁移
迁移规划
总迁移时间
software-defined networking(SDN)
data center
virtual machine migration
migration plan
total migration time
